<br/>A novel approach to accelerate the development of coarse-grained (CG) models is used by integrating dynamic artificial neural network (DANN), particle swarm optimization (PSO), and molecular dynamics (MD) simulations has been proposed. CG models of two solvents, namely, D2O and Dimethyl fumarate (DMF) have been developed at 300 K. We show that the DANN model trained on the data obtained from PSO and MD simulations can accelerate the development of CG models. The new models can predict the structural and dynamical properties of both D2O and DMF with precision under different conditions. This new method has tremendous promise in accelerating the atomic, coarse-grained, and reactive models of different materials and solvents.<br/>

Machine learnt meso-scale models

255th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.